| CPC G06N 20/00 (2019.01) [G06F 17/18 (2013.01); G06N 5/04 (2013.01); G06Q 30/0275 (2013.01)] | 20 Claims |
|
1. A method, comprising:
generating, using a first loss function, a plurality of loss values associated with a plurality of sets of auction information, wherein a first loss value of the plurality of loss values is associated with a first set of auction information associated with a first auction;
training a machine learning model using the first loss function, the plurality of loss values and the plurality of sets of auction information to generate a first machine learning model comprising a plurality of feature parameters associated with a plurality of features of the plurality of sets of auction information, wherein:
the first loss function comprises a first value and a second value;
the first value corresponds to:
a first minimum bid value to win the first auction; or
an optimal bid reduction factor determined based upon the first minimum bid value to win the first auction and a first bid value associated with a first content item;
the second value corresponds to:
a first shaded bid value associated with the first content item; or
a bid reduction factor used to determine the first shaded bid value;
the generating the plurality of loss values comprises generating the first loss value based upon a difference between the first value and the second value; and
the first machine learning model is generated using a plurality of win-rates comprising a first win-rate corresponding to a first quantity of won auctions associated with the first shaded bid value;
loading the first machine learning model onto a bid shading module of a demand-side platform (DSP), wherein the DSP is at least partially implemented by a DSP server;
receiving, by the DSP at least partially implemented by the DSP server, a bid request from at least one of a supply-side platform (SSP) server or a content exchange server, wherein:
the bid request is associated with a request for content associated with a client device; and
the bid request is indicative of a set of features comprising one or more features associated with the request for content;
determining a second bid value associated with a second content item;
inputting, into the bid shading module of the DSP at least partially implemented by the DSP server, the second bid value and one or more first feature parameters, of the plurality of feature parameters, associated with the set of features;
determining, using the first machine learning model loaded onto the bid shading module of the DSP at least partially implemented by the DSP server, a second shaded bid value associated with the second content item based upon the second bid value and the one or more first feature parameters, of the plurality of feature parameters, associated with the set of features; and
submitting the second shaded bid value to an auction module that is at least partially implemented by at least one of the SSP server or the content exchange server for participation in a second auction associated with the request for content,
wherein the second content item is provided for presentation on the client device associated with the request for content based upon a determination that the second content item is a winner of the second auction, wherein use of the first machine learning model that was generated using the plurality of win-rates increases a total win-rate associated with a plurality of auctions.
|